CHEMICAL ANALYSIS OF WHEAT-FLOUR SUBSTITUTES. 3 
until thoroughly dry and brittle, and then passing the dried product 
through a small hand-power mill. For the banana flour, large 
green fruit was used. It was peeled by hand, sliced, dried, and 
ground into a flour. 
Of those prepared in the Plant Chemical Laboratory the chestnut, 
dasheen (peeled and unpeeled), rice (polished and brown), millet, 
bean, pea, chick-pea, and soy bean were first thoroughly cleansed and 
then ground in a hand mill, while rye, feterita, kafir, milo, and 
kaoliang were scoured and then ground in an experimental mill 
to give the desired flours. To obtain the peanut meal, roasted 
peanuts were crushed with a rolling pin. 
The potato flakes employed in these experiments were obtained 
from Germany several years ago. The boiled potatoes were prepared 
in the laboratory by boiling potatoes in their skins, after which 
they were cooled, peeled, and mashed. 
BREAD. 
All the breads analyzed in this investigation were made from 
mixtures of flour, in the proportion of 25 per cent of substitute to 
75 per cent of standard wheat flour. 1 
METHODS OF ANALYSIS. 
The methods for the analysis of cereal foods adopted by the 
Association of Official Agricultural Chemists 2 were used for all the 
determinations made, with the exception of the fat determination. 
As it is impossible to obtain accurate results for the estimation of 
fat in bread by the usual method of extraction, the Polenski method 3 
was employed for this determination. The carbohydrates were 
obtained by difference. The calories per pound were calculated by 
multiplying the sum of the carbohydrates and protein by 1,860, and 
the fat by 4,220, and then adding the two results thus obtained. 4 
RESULTS OF THE ANALYSIS. 
Table 1 gives the composition of the standard flour and of the 
substitutes. The composition of the breads, as analyzed on the 
air-dry basis, is given in Table 2, and, as calculated on the basis of 
an assumed average moisture content of 35 per cent, in Table 3. In 
a general way, the data in these three tables have been arranged in 
the order of the protein content of the flour. Those showing a low 
protein content come first, being followed in turn by those pre- 
pared from grains and cereals, then by those made from legumes, 
and, finally, by those from certain by-products. Besides giving 
the composition of the various flours and breads studied, the tables 
i The bread was made according to the directions given in U. S. Dept. Agr. Farmers' Bui. 955. 
2 Jour. A. O. A. C, Aug. 15, 1916, Vol. II, No. 2, part 2: 187. 
3 Zt. Anal. Chem. 1911, 50: 655. 
4 The calories were calculated by Rubner's factors. 
